Meta has enabled voice conversations with its Muse AI agent on iOS, allowing users to speak directly to the assistant while observing an animated avatar that reacts to the dialogue. CEO Mark Zuckerberg confirmed the update today, noting that the feature permits real-time interaction where the character’s expressions and movements synchronise with the generated speech. Users can adjust the vocal tone through specific settings or instruct the agent to select an appropriate style automatically. Testing indicates the system responds quickly to voice commands, though the agent occasionally pauses to retrieve information before replying. The interface displays the avatar at the top of the chat window, ensuring visual feedback matches the audio output during the session.
This shift marks a practical step toward making the agent feel more present during interactions, moving beyond text-only exchanges to include auditory and visual cues. The capability allows for a more natural conversational flow without altering the underlying model training or API specifications. It demonstrates how Meta is integrating multimodal feedback into existing mobile applications.
* The update is currently available on the iOS app.
* Voice responses occur in real time with occasional pauses for data retrieval.
* Users can manually or automatically adjust the avatar’s vocal tone.
